为智能业务运营编排代理AI

How IBM can help IBM brings deep business process outsourcing industry expertise, partnerships, andexperience to transform intelligent operational workflows, helping businesses unlockvalue with AI agents and assets. For more information, visithttps://www.ibm.com/consulting/operations People remainessential to agentic AI. AI agents accelerateoperational velocity. AI agentselevate employees. 90% 75% 81% of executives say that by 2027, AIagents will enable business operationsprofessionals to move beyond simplereporting to perform insightful analyticsfor real-time optimization. of executives say AI agents will executetransactional processes and workflowsautonomously within two years for24x7 availability. of executives recognize that to seizecompetitive advantage, they need theright people in the right positions withthe right incentives. AI agents changethe playing field and communication. Human oversight andconnections, decision-making, and mostcritical, creativity, are more importantthan ever in this next generation ofintelligent operations. In this research brief, we review threeelements of autonomous automation.Part one explores a new workplaceparadigm where technology acts whilepeople orchestrate. In part two, wediscuss how AI agents focus on outcomes,not rules, to improve results, while peopleadd value throughout. Part threeexamines how best to scale automationacross global business operations, withexpert strategists averting potentialpitfalls. The report concludes with anaction guide that outlines specific stepsyou can take today to capture theadvantages of autonomous automation. Automation is a journey (see Perspective,“Automation progression” on page 6). Asorganizations move from basicautomation of tasks, to automatinginterconnected workflows, to fullyautonomous systems, they can designa new operating model for businessoperations (see Figure 1). As with eachform of AI-enabled automation, theeffectiveness of this model dependson the data fueling agents and thegovernance managing outcomes. Organizations are turning to agentic AI—AI systems that can accomplish specificgoals without supervision1—to capturea competitive edge. With agentic AI, operationsautonomously learn, adapt, and optimizein real time. It isn’t just about being fasterand more efficient. It’s about intelligentagents that proactively anticipatechallenges, personalize experiences,and drive innovation. It’s a shift fromautomating tasks to orchestratingever-adapting processes. New research from the IBM Institute forBusiness Value (IBM IBV) highlights thedramatic evolution underway. Wesurveyed 750 cross-industry operationsexecutives across six countries. More than80% cite automating global businessservices as a major strategic imperative.And they expect AI agents to get themthere. 86% say that by 2027, processautomation and workflow reinvention willbe more effective because of AI agents. But successful deployment of agents isn’tjust about the tech. It depends on creatingsynergy between people and AI acrossvirtually every operational transaction Figure 1 In an agentic AI operating model, agents and people work in tandem,using the inputs on the left in a continuous flow of interaction, reaction,decision, and action to enhance business operations. Ecosystem partner systems Order-to-cash workflow automation Sales support & customer service External data Sales forecasting, lead prioritization, and personalizedmarketing from real-time customer and market insights. Customer assistance with 24x7 multilingual globalsupport; customized and proactive responses. Perspective AI automation journeys begin with rules-based systems that handle repetitivetasks, evolving to AI assistants that helpwith tasks through natural languageinteractions, and finally to AI agents,which are broader systems that carry outtasks and make decisions autonomouslywith little human input (see figure). AI assistants and AI agents play different roles. Here’s what sets them apart:2 Automationprogression Assistants Agents –Perform tasks as requestedthrough prompts–Use a conversational AI interface–Base insights on machine learningor foundation models –Work autonomously to achievespecific goals and KPIs–Act independently, deciding whichtools to use and when–Use persistent memory and adaptivelearning to refine their approach AI benefits business operations at eachmaturity stage, but agentic AI holdstransformational potential. Agentic AIand autonomy Interconnected andintelligent workflows AI-enabledautomation Basicautomation Orchestrate and executeworkflows autonomously Type of impact Automate high-volume,repetitive tasks Scale expertise and reachwith AI assistants Augment workforcecapabilities Use AI agents to continuouslyimprove business operationsby making feedback-basedadjustments—autonomously—with touchless 24x7 operations Integrate workflows withAI assistants to providereal-time and personalizedresponses to customer,employee, and partneri